Visual Servoing and Pose Estimation with Cameras Obeying the Unified Model - Visual Servoing via Advanced Numerical Methods - page 239

13.8(b) give the obtained results with 0 3 pixels gaussian noises, principal points co-

ordinates [375 375] and focal F = 760 (recall that the real values are [400 400] for
the principal point and F = 800 for the focal). From these figures, it can be noticed
that the accuracy of the estimation using the linear method decreases when the data
noise increases. The results obtained using the linear method are improved using
Araujo's method. On the other hand, the accuracy of the Araujo iterative method
initialized by the identity matrix also decreases, but the convergence percentage is
still around 40%. Finally for the same experiment, our iterative method converges
for all cases and gives more accurate estimation of the poses.

.

13.6 Conclusion

In this chapter, a unique and efficient decoupled scheme for visual servoing and pose
estimation has been proposed. The proposed scheme is valid for cameras obeying
the unified model. More precisely, the invariants to rotational motions computed
from the projection onto the unit sphere are used to control the translational DOF.
Adequate forms of invariants have been proposed to decrease the interaction matrix
variations with respect to the depth distributions.The validations results have shown
the efficiency of the proposed scheme. Future works will be devoted to extend these
results to model-free pose estimation problem and to region-based visual servoing.

References

[1] Ansar, A., Daniilidis, K.: Linear pose estimation from points or lines. IEEE Trans. on
Pattern Analysis and Machine Intelligence 25, 282-296 (2003)
[2] Araujo, H., Carceroni, R.L., Brown, C.M.: A fully projective formulation to improve
the accuracy of lowe's pose-estimation algorithm. Computer Vision and Image Under-
standing 70, 227-238 (1998)
[3] Baker, S., Nayar, S.: A theory of catadioptric image formation. Int. Journal of Computer
Vision 35(2), 175-196 (1999)
[4] Barreto, J., Araujo, H.: Issues on the geometry of central catadioptric image formation.
In: Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision
and Pattern Recognition. CVPR 2001, vol. 2, pp. II-422-II-427 (2001)
[5] Chaumette, F.: Potential problems of stability and convergence in image-based and
position-based visual servoing. In: Kriegman, D., Hager, G., Morse, A. (eds.) The Con-
fluence of Vision and Control. LNCIS, vol. 237, pp. 66-78. Springer, Heidelberg (1998)
[6] Corke, P.I., Hutchinson, S.A.: A new partitioned approach to image-based visual servo
control. IEEE Trans. on Robotics and Automation 17(4), 507-515 (2001)
[7] Dementhon, D., Davis, L.: Model-based object pose in 25 lines of code. Int. Journal of
Computer Vision 15(1-2), 123-141 (1995)
[8] Dhome, M., Richetin, M., Lapreste, J.T., Rives, G.: Determination of the attitude of 3d
objects from a single perspective view. IEEE Trans. on Pattern Analysis and Machine
Intelligence 11(12), 1265-1278 (1989)
[9] Espiau, B., Chaumette, F., Rives, P.: A new approach to visual servoing in robotics.
IEEE Trans. on Robotics and Automation 8, 313-326 (1992)